Machine Learning Approaches for Quality Assessment of Protein Structures

Overview of AlphaFold2 and breakthroughs in overcoming its limitations

Predicting three-dimensional (3D) protein structures has been challenging for decades. The emergence of AlphaFold2 (AF2), a deep learning-based machine learning method developed by DeepMind, became a game changer in the protein folding community. AF2 can predict a protein's three-dimensional structure with high confidence based on its amino acid sequence. Accurate prediction of protein structures can dramatically accelerate our understanding of biological mechanisms and provide a solid foundation for reliable drug design. Although AF2 breaks through the barriers in predicting protein structures, many rooms remain to be further studied. This review provides a brief historical overview of the development of protein structure prediction, covering template-based, template-free, and machine learning-based methods. In addition to reviewing the potential benefits (Pros) and considerations (Cons) of using AF2, this review summarizes the diverse applications, including protein structure predictions, dynamic changes, point mutation, integration of language model and experimental data, protein complex, and protein-peptide interaction. It underscores recent advancements in efficiency, reliability, and broad application of AF2. This comprehensive review offers valuable insights into the applications of AF2 and AF2-inspired AI methods in structural biology and its potential for clinically significant drug target discovery.

Computational resources and chemoinformatics for translational health research

The integration of computational resources and chemoinformatics has revolutionized translational health research. It has offered a powerful set of tools for accelerating drug discovery. This chapter overviews the computational resources and chemoinformatics methods used in translational health research. The resources and methods can be used to analyze large datasets, identify potential drug candidates, predict drug-target interactions, and optimize treatment regimens. These resources have the potential to transform the drug discovery process and foster personalized medicine research. We discuss insights into their various applications in translational health and emphasize the need for addressing challenges, promoting collaboration, and advancing the field to fully realize the potential of these tools in transforming healthcare.

High-accuracy protein model quality assessment using attention graph neural networks

Great improvement has been brought to protein tertiary structure prediction through deep learning. It is important but very challenging to accurately rank and score decoy structures predicted by different models. CASP14 results show that existing quality assessment (QA) approaches lag behind the development of protein structure prediction methods, where almost all existing QA models degrade in accuracy when the target is a decoy of high quality. How to give an accurate assessment to high-accuracy decoys is particularly useful with the available of accurate structure prediction methods. Here we propose a fast and effective single-model QA method, QATEN, which can evaluate decoys only by their topological characteristics and atomic types. Our model uses graph neural networks and attention mechanisms to evaluate global and amino acid level scores, and uses specific loss functions to constrain the network to focus more on high-precision decoys and protein domains. On the CASP14 evaluation decoys, QATEN performs better than other QA models under all correlation coefficients when targeting average LDDT. QATEN shows promising performance when considering only high-accuracy decoys. Compared to the embedded evaluation modules of predicted ${C}_{\alpha^{-}} RMSD$ (pRMSD) in RosettaFold and predicted LDDT (pLDDT) in AlphaFold2, QATEN is complementary and capable of achieving better evaluation on some decoy structures generated by AlphaFold2 and RosettaFold. These results suggest that the new QATEN approach can be used as a reliable independent assessment algorithm for high-accuracy protein structure decoys.

CNN-Pred: Prediction of single-stranded and double-stranded DNA-binding protein using convolutional neural networks

DNA-binding proteins play a vital role in biological activity including DNA replication, DNA packing, and DNA reparation. DNA-binding proteins can be classified into single-stranded DNA-binding proteins (SSBs) or double-stranded DNA-binding proteins (DSBs). Determining whether a protein is DSB or SSB helps determine the protein's function. Therefore, many studies have been conducted to accurately identify DSB and SSB in recent years. Despite all the efforts have been made so far, the DSB and SSB prediction performance remains limited. In this study, we propose a new method called CNN-Pred to accurately predict DSB and SSB. To build CNN-Pred, we first extract evolutionary-based features in the form of mono-gram and bi-gram profiles using position specific scoring matrix (PSSM). We then, use 1D-convolutional neural network (CNN) as the classifier to our extracted features. Our results demonstrate that CNN-Pred can enhance the DSB and SSB prediction accuracies by more than 4%, on the independent test compared to previous studies found in the literature. CNN-pred as a standalone tool and all its source codes are publicly available at: https://github.com/MLBC-lab/CNN-Pred.

A Methylation Diagnostic Model Based on Random Forests and Neural Networks for Asthma Identification

Background

Asthma significantly impacts human life and health as a chronic disease. Traditional treatments for asthma have several limitations. Artificial intelligence aids in cancer treatment and may also accelerate our understanding of asthma mechanisms. We aimed to develop a new clinical diagnosis model for asthma using artificial neural networks (ANN).

Methods

Datasets (GSE85566, GSE40576, and GSE13716) were downloaded from Gene Expression Omnibus (GEO) and identified differentially expressed CpGs (DECs) enriched by Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis. Random forest (RF) and ANN algorithms further identified gene characteristics and built clinical models. In addition, two external validation datasets (GSE40576 and GSE137716) were used to validate the diagnostic ability of the model.

Results

The methylation analysis tool (ChAMP) considered DECs that were up-regulated (n =121) and down-regulated (n =20). GO results showed enrichment of actin cytoskeleton organization and cell-substrate adhesion, shigellosis, and serotonergic synapses. RF (random forest) analysis identified 10 crucial DECs (cg05075579, cg20434422, cg03907390, cg00712106, cg05696969, cg22862094, cg11733958, cg00328720, and cg13570822). ANN constructed the clinical model according to 10 DECs. In two external validation datasets (GSE40576 and GSE137716), the Area Under Curve (AUC) for GSE137716 was 1.000, and AUC for GSE40576 was 0.950, confirming the reliability of the model.

Conclusion

Our findings provide new methylation markers and clinical diagnostic models for asthma diagnosis and treatment.

Machine learning approaches and their applications in drug discovery and design

This review is focused on several machine learning approaches used in chemoinformatics. Machine learning approaches provide tools and algorithms to improve drug discovery. Many physicochemical properties of drugs like toxicity, absorption, drug-drug interaction, carcinogenesis, and distribution have been effectively modeled by QSAR techniques. Machine learning is a subset of artificial intelligence, and this technique has shown tremendous potential in the field of drug discovery. Techniques discussed in this review are capable of modeling non-linear datasets, as well as big data of increasing depth and complexity. Various machine learning-based approaches are being used for drug target prediction, modeling the structure of drug target, binding site prediction, ligand-based similarity searching, de novo designing of ligands with desired properties, developing scoring functions for molecular docking, building QSAR model for biological activity prediction, and prediction of pharmacokinetic and pharmacodynamic properties of ligands. In recent years, these predictive tools and models have achieved good accuracy. By the use of more related input data, relevant parameters, and appropriate algorithms, the accuracy of these predictions can be further improved.

An Overview of Alphafold's Breakthrough

This paper presents a short summary of the protein folding problem, what it is and why it is significant. Introduces the CASP competition and how accuracy is measured. Looks at different approaches for solving the problem followed by a review of the current breakthroughs in the field introduced by AlphaFold 1 and AlphaFold 2.

Chemical toxicity prediction based on semi-supervised learning and graph convolutional neural network

As safety is one of the most important properties of drugs, chemical toxicology prediction has received increasing attentions in the drug discovery research. Traditionally, researchers rely on in vitro and in vivo experiments to test the toxicity of chemical compounds. However, not only are these experiments time consuming and costly, but experiments that involve animal testing are increasingly subject to ethical concerns. While traditional machine learning (ML) methods have been used in the field with some success, the limited availability of annotated toxicity data is the major hurdle for further improving model performance. Inspired by the success of semi-supervised learning (SSL) algorithms, we propose a Graph Convolution Neural Network (GCN) to predict chemical toxicity and trained the network by the Mean Teacher (MT) SSL algorithm. Using the Tox21 data, our optimal SSL-GCN models for predicting the twelve toxicological endpoints achieve an average ROC-AUC score of 0.757 in the test set, which is a 6% improvement over GCN models trained by supervised learning and conventional ML methods. Our SSL-GCN models also exhibit superior performance when compared to models constructed using the built-in DeepChem ML methods. This study demonstrates that SSL can increase the prediction power of models by learning from unannotated data. The optimal unannotated to annotated data ratio ranges between 1:1 and 4:1. This study demonstrates the success of SSL in chemical toxicity prediction; the same technique is expected to be beneficial to other chemical property prediction tasks by utilizing existing large chemical databases. Our optimal model SSL-GCN is hosted on an online server accessible through: https://app.cbbio.online/ssl-gcn/home .

Deep Learning-Based Advances in Protein Structure Prediction

Obtaining an accurate description of protein structure is a fundamental step toward understanding the underpinning of biology. Although recent advances in experimental approaches have greatly enhanced our capabilities to experimentally determine protein structures, the gap between the number of protein sequences and known protein structures is ever increasing. Computational protein structure prediction is one of the ways to fill this gap. Recently, the protein structure prediction field has witnessed a lot of advances due to Deep Learning (DL)-based approaches as evidenced by the success of AlphaFold2 in the most recent Critical Assessment of protein Structure Prediction (CASP14). In this article, we highlight important milestones and progresses in the field of protein structure prediction due to DL-based methods as observed in CASP experiments. We describe advances in various steps of protein structure prediction pipeline viz. protein contact map prediction, protein distogram prediction, protein real-valued distance prediction, and Quality Assessment/refinement. We also highlight some end-to-end DL-based approaches for protein structure prediction approaches. Additionally, as there have been some recent DL-based advances in protein structure determination using Cryo-Electron (Cryo-EM) microscopy based, we also highlight some of the important progress in the field. Finally, we provide an outlook and possible future research directions for DL-based approaches in the protein structure prediction arena.

QUARTERplus: Accurate disorder predictions integrated with interpretable residue-level quality assessment scores

A recent advance in the disorder prediction field is the development of the quality assessment (QA) scores. QA scores complement the propensities produced by the disorder predictors by identifying regions where these predictions are more likely to be correct. We develop, empirically test and release a new QA tool, QUARTERplus, that addresses several key drawbacks of the current QA method, QUARTER. QUARTERplus is the first solution that utilizes QA scores and the associated input disorder predictions to produce very accurate disorder predictions with the help of a modern deep learning meta-model. The deep neural network utilizes the QA scores to identify and fix the regions where the original/input disorder predictions are poor. More importantly, the accurate QUATERplus's predictions are accompanied by easy to interpret residue-level QA scores that reliably quantify their residue-level predictive quality. We provide these interpretable QA scores for QUARTERplus and 10 other popular disorder predictors. Empirical tests on a large and independent (low similarity) test dataset show that QUARTERplus predictions secure AUC = 0.93 and are statistically more accurate than the results of twelve state-of-the-art disorder predictors. We also demonstrate that the new QA scores produced by QUARTERplus are highly correlated with the actual predictive quality and that they can be effectively used to identify regions of correct disorder predictions. This feature empowers the users to easily identify which parts of the predictions generated by the modern disorder predictors are more trustworthy. QUARTERplus is available as a convenient webserver at http://biomine.cs.vcu.edu/servers/QUARTERplus/.

In Silico Tools and Approaches for the Prediction of Functional and Structural Effects of Single-Nucleotide Polymorphisms on Proteins: An Expert Review

Single-nucleotide polymorphisms (SNPs) are single-base variants that contribute to human biological variation and pathogenesis of many human diseases. Among all SNP types, nonsynonymous single-nucleotide polymorphisms (nsSNPs) can alter many structural, biochemical, and functional features of a protein such as folding characteristics, charge distribution, stability, dynamics, and interactions with other proteins/nucleotides. These modifications in the protein structure can lead nsSNPs to be closely associated with many multifactorial diseases such as cancer, diabetes, and neurodegenerative diseases. Predicting structural and functional effects of nsSNPs with experimental approaches can be time-consuming and costly; hence, computational prediction tools and algorithms are being widely and increasingly utilized in biology and medical research. This expert review examines the in silico tools and algorithms for the prediction of functional or structural effects of SNP variants, in addition to the description of the phenotypic effects of nsSNPs on protein structure, association between pathogenicity of variants, and functional or structural features of disease-associated variants. Finally, case studies investigating the functional and structural effects of nsSNPs on selected protein structures are highlighted. We conclude that creating a consistent workflow with a combination of in silico approaches or tools should be considered to increase the performance, accuracy, and precision of the biological and clinical predictions made in silico.